Account

What new tool do you want?

Math & Education

Rounding & Significant Figures Calculator

Round a number to any decimal place or number of significant figures — half up, half even (banker's), always up or always down — and count its sig figs.

How to use

  1. 1Enter a number and choose decimal places or significant figures.
  2. 2Pick the rounding rule.

How it's calculated

Half up rounds 0.5 away from zero; half even (banker's rounding) rounds 0.5 to the nearest even digit. Sig-fig counting: leading zeros never count; trailing zeros count only after a decimal point.

Frequently asked questions

How many significant figures does 0.00520 have?

Three — the leading zeros don't count, but the trailing zero after the decimal point does.

What is banker's rounding?

Rounding exact halves to the nearest even number (2.5 → 2, 3.5 → 4), which avoids an upward bias in sums. It's the IEEE 754 default.

How do I round to the nearest ten or hundred?

Use negative decimal places: −1 rounds to the nearest 10, −2 to the nearest 100.

Rounding to significant figures first converts the request into decimal places

Internally, 'round to 3 significant figures' is translated into an equivalent decimal-places instruction before the actual rounding happens: the calculator finds the order of magnitude of the number (via log₁₀) and computes how many places after the decimal point that many significant figures would require. For 0.0052049 rounded to 3 significant figures, that works out to 5 decimal places, giving 0.00520 — the same underlying roundTo() routine handles both modes, with significant-figure rounding simply a smarter way of deciding how many decimal places to ask for based on where the number's first nonzero digit falls.

Worked example: why 0.00520 has 3 significant figures, not 5

Leading zeros (the three before the 5) never count — they only mark the decimal position, not measured precision. But trailing zeros after a decimal point do count, because writing "0.00520" instead of "0.0052" is a deliberate signal that the measurement is precise to that extra zero. So the significant figures here are 5, 2 and 0 — three total, which is exactly what this tool reports for that input.

Why 2.5 can round to 2 under one rule and 3 under another

Standard "half up" rounding always pushes an exact .5 away from zero, so 2.5 → 3. Banker's rounding (half-to-even) instead rounds to whichever neighbor is even, so 2.5 → 2 but 3.5 → 4. Banker's rounding exists because always rounding .5 upward introduces a small systematic upward bias when you round and then sum many numbers; alternating between even neighbors cancels that bias out on average, which is why it's the default behavior in IEEE 754 floating-point arithmetic used by most programming languages.

Negative decimal places round to tens, hundreds, and beyond

Rounding 1234.5 to -2 decimal places means rounding to the nearest hundred: the calculator scales by 10⁻² first, rounds, then scales back, landing on 1200. That's the same operation as rounding to 2 significant figures would be for a number in that range, just expressed through place value instead of digit count — useful for things like rounding a budget estimate to the nearest thousand rather than specifying an arbitrary number of significant digits.

'Half up' here means away from zero, not toward positive infinity

Rounding -2.5 to the nearest whole number under this tool's 'half up' rule gives -3, not -2 — the rounding direction is defined relative to zero (push the halfway case further from it), matching how most people round by hand, rather than always rounding toward positive infinity the way some spreadsheet functions do. That distinction only shows up on negative halfway values, but it's exactly the kind of mismatch that makes a rounded result disagree between two different tools or programming languages even though both claim to do ordinary 'round half up' rounding.

Related tools

More: Math & Education